Environ. Sci. Technol. 1092, 26, 352-360
Carlberg, G. E.; Martinsen, K.; Kringstad, A,; Gjessing, E.; Grande, M.; Kiillqvist, T.; Skare, J. U. Arch. Enuiron. Contam. Toxicol. 1986,15, 543-548. Casselman, J. M. Calcified tissue and body growth of northern pike, Eson lucius Linneus. Doctoral Dissertation. University of Toronto, Toronto, Canada, 1978. Okla, L.; WesBn, C. J. Chromatogr. 1984,299,420-423. Duinker, J. C.; Hillebrand, M. T. J. Bull. Enuiron. Contam. Toxicol. 1983,31,25-32. Ballschmiter, K.;Zell, M. Fresenius 2.Anal. Chem. 1980, 302,20-31. Newton, I. Enuiron. Pollut. 1988,55,29-40. Swedish Standards for Environmental Analysis, Svensk Standard, SIS, 1991.
(27) Jensen, A. L.; Spigarelli, S. A.; Thommes, M. M. Can. J. Fish. Aquat. Sci. 1982,39,700-709. (28) Borgmann, U. Sci. Total Enuiron. 1985,44, 111-121. (29) Black, M. C.; McCarthy, J. F. Enuiron. Toxicol. Chem. 1988, 7,593-600. (30) McCarthy, J. F.; Jimenez, B. D. I. Enuiron. Sci. Technol. 1985.19. 1072-1075. (31) C&r, C. W.; Suffet, I. H. Enuiron. Sci. Technol. 1982,16, 735-1075. (32) Rice, C.; White, D. Enuiron. Toxicol. Chem. 1987, 6, 259-274.
Received for review July 10,1991.Accepted September 18,1991.
Differential Transport of Sewage-Derived Nitrogen and Phosphorus through a Coastal Watershed Peter K. Weiskei” Geology Department, Boston University, Boston, Massachusetts 022 15
Brian L. Howes Biology Department, Woods Hole Oceanographic Institution, Woods Hole, Massachusetts 02543
Changes of land use in coastal watersheds to residential development with on-site sewage disposal represent a potential change in both the quantity and quality of nutrient inputs to coastal marine systems. Measurements of dissolved N and phosphate P in septic system effluent indicated initial concentrations 100-1000-fold greater than receiving coastal waters, with inorganic N / P ratios (17/1) similar to phytoplankton growth requirements. Transformations of organic and inorganic N and retention of inorganic P occurred in the initial meters of groundwater transport with substantial ( ~ 7 0 % nitrification ) of effluent ammonium to nitrate and retention of phosphate by the soil ( ~ 6 0 % ) .The degree of initial transformation and retention was directly related to unsaturated infiltration distance and is consistent with the requirements of these processes for oxidizing conditions. At greater distances (10-100 m), over 99% of the total dissolved N occurred as nitrate, phosphate concentrations were reduced to background levels, and groundwater N / P ratios exceeded 2500/1. The greater the importance of high-N, low-P groundwater inputs to the nutrient balance of a coastal water body, the greater the potential for shifts in the nutrient which limits primary production.
Introduction Each year in the United States, approximately 3.8 billion cubic meters of sewage effluent is discharged to the subsurface by on-site septic systems (I);an increasing fraction of this total is discharged to coastal watersheds (2). Septic effluent contains a variety of microbial, inorganic, and organic constituents (3-5) and is highly enriched in nitrogen and phosphorus compared to most groundwaters and surface waters. Dissolved N and P concentrations in domestic effluent are on the order of 3000 and 400 pM, respectively (4, 6 ) , which are 100-1000-fold higher than concentrations in typical receiving water bodies. The fate of this anthropogenic contribution is of both ecological and management importance, because of the key role played by N and P in structuring aquatic ecosystems. Field studies have shown that primary production in freshwater ecosystems is generally P-limited (7-9). Coastal marine systems, by contrast, are more often N-limited (10, 11),while estuaries show great variation in nutrient lim352
Environ. Sci. Technol., Vol. 26, No. 2, 1992
itation (12).One factor controlling the process of nutrient enrichment or eutrophication in these ecosystems is the rate at which “new” N or P is added from terrestrial sources. In lakes, embayments, and estuaries where terrestrial nutrient inputs derive mainly from streamflow and point discharges, N and P transport is frequently modified by plant uptake and generally well-studied (12-14). In permeable, groundwater-dominated watersheds, where nutrients are injected to the subsurface by sewage disposal or other activities, the total mass flux and subsurface behavior of N and P are poorly understood and appear to be controlled by bacterial and chemical interactions (15-17). While simple loading models are available for estimation of nutrient inputs to such watersheds (18), assumptions regarding subsurface behavior are usually necessary in order to quantify watershed nutrient outputs. In the present study, we followed the transport of effluent-derived, plant-available inorganic N and P through the freshwater portion of a Massachusetts coastal watershed. Our goal was to determine both the overall nutrient retention and the change in the ratio of N to P during groundwater transport, since both factors are central to understanding the potential impact of this nutrient source on coastal waters. It should be noted that all studies of septic system impact must address a difficult problem of scale. Most previous workers have concentrated either on the microscale and mapped individual septic system plumes in detail (19,20)or on the regional scale and examined the relationship between residential land use and groundwater quality across entire basins (21-25). To evaluate thoroughly the fate of on-site sewage inputs in this study area, it was necessary to use a combination of both approaches, since our purpose was both to characterize the source and to quantify the overall extent of N and P transport through the watershed.
Study Area The Indian Heights subbasin is located adjacent to Buttermilk Bay, at the northern tip of Buzzards Bay in southeast Massachusetts (Figure 1). The regional geology is dominated by medium-to-coarse sands, which were deposited in a 200-km2outwash plain during the retreat of
0013-936X/92/0926-0352$03.00/0
0 1992 American Chemical Society
magnesian minerals and metal oxides, and very little clay [generally 0.10), while the PO4 loss was approximately 60%. Near-field phosphate losses caused a sharp increase in the median molar ratio of DIN to P, from 17 in the effluent to 69 in the near-field groundwater. Near-field DIN/P ratios were found to vary exponentially with mean infiltration distance (the distance between the base of the septic absorption system and the water table; Figure 4). Significant changes also occurred in the relative proportions of dissolved N species during early transport.
4
.+ _
9 =
3
.-
-
\
x = infiltration distance
y = N/P ratio
4
\
0.8514
0
\
&I
0
Flgure 4. Near-field groundwater DIN/P ratio versus inflitrationdistance at the time of sampling, sites 1-4. Infittration distance Is the distance from base of leaching field or cesspool to the water table. Data from four dates, December 1987-June 1988.
While over 99% of the effluent DIN was ammonium, only 27% of the near-field groundwater DIN remained in this form, presumably due to the rapid conversion of NH4+to NO3- by nitrifying bacteria. As with phosphate removal, the extent of nitrification was related to mean infiltration distance; nitrate fractions of DIN in near-field groundwater ranged from a low of 38% (site 4, with zero infiltration distance) to a high of 99% (site 2, with 3.5-m infiltration distance). Dissolved organic N (DON, obtained by difference from TDN and DIN data) averaged 499 f 86 pM in the effluent samples (mean f SEM) and comprised 16% of the TDN in the effluent (Table I). As with phosphate removal and nitrification, the degree of DON removal and/or mineralization during early transport was strongly related to infiltration distance. DON was below detection in 60% of the near-field groundwater samples and was only detected at site 4,where the mean concentration was 468 f 378 pM and the infiltration distance was zero. The increased phosphate removal, nitrification, and DON mineralization associated with increased infiltration distance suggests that these processes either require, or are strongly enhanced by, oxidizing conditions and/or free molecular oxygen. In the sandy soils of this study area, the likelihood of finding these conditions beneath an operating septic system is controlled, at least in part, by infiltration distance. Vertical profiles obtained 5 m downgradient of septic system 1confirmed the pattern of nutrient behavior described above. Specific conductance and nitrate concentration were well correlated over the upper four sampling levels (n = 16; r2 = 0.88; Figure 5A,B), with both conductance and nitrate profiles showing peak values at the water table. [The conductance peak at the 2.35-m depth is due to a tongue of saline groundwater from Buttermilk Bay; mean high water is 9 m downgradient of the cross section and the conductance observed (14 800 pS cm-l) is >10 times higher than the measured effluent conductance and about half the conductance of seawater.] Peak ammonium and phosphate concentrations, by contrast, were found not at the water table, but -35 cm below it (Figure 5C,D). The depressed ammonium and phosphate concentrations at the water table are probably not due to dilution with natural recharge, since the nitrate and conductance profiles are not similarly affected. Rather, ammonium transformation and phosphate removal appear to be enhanced by redox conditions at the water table, which are likely to be oxidizing relative to the plume core. The conductance, nitrate, and ammonium profiles decline Envlron. Sci. Technol., Vol. 26, No. 2. 1992
355
(uhl)
PG4-P
0
26
50
40
t
W
2
1 c
cl 0
3
3
4i
D
3
4Flgure 5. Near-field groundwater profiles, 5 m downgradient, site 1, September 1988. Mean at each depth, averaged across four sampling points in transverse cross section: (A) specific conductance, (B) nltrate nitrite, (C) ammonium, (D) phosphate. Error bars = SEM.
+
Table 11. Watershed N u t r i e n t Concentrations and Specific Conductance (Mean f S E M )
well or sampler
spec conduct., pS/cm
NH4+pM
NO2- + NO3- pM
Po43-,pM
DIN, pM
Upgradient Wells" w-5 W-6
40 34
0.7 0.7
w-1 w-11 w-12
270 188 104 114
